Article pubs.acs.org/cm
Antifibrinogen, Antireflective, Antifogging Surfaces with Biocompatible Nano-Ordered Hierarchical Texture Fabricated by Layer-by-Layer Self-Assembly Kengo Manabe,† Motomi Matsuda,† Chiaki Nakamura,† Keisuke Takahashi,† Kyu-Hong Kyung,‡ and Seimei Shiratori*,† †
Center for Material Design Science, School of Integrated Design Engineering, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa 223-8522, Japan ‡ SNT Company, Ltd., 7-1 Shinkawasaki, Saiwai-ku, Kawasaki, Kanagawa 212-0032, Japan S Supporting Information *
ABSTRACT: Endoscopic surgery is a minimally invasive approach that is widely used in various clinical departments, including digestive surgery, thoracic surgery, and urology, because it can minimize the burden on patients. To perform more elaborate procedures, highly functional coatings that enhance the operation efficiency of the related equipment are required; for example, coatings to improve the visibility through endoscope lenses are needed. In this study, we designed multifunctional surfaces that displayed antithrombogenicity, antireflection, and antifogging by controlling nano-ordered hierarchical structures fabricated via layer-by-layer self-assembly. The coatings were composed of polyelectrolyte multilayers prepared from blends of poly(vinyl alcohol) (PVA) and poly(acrylic acid) (PAA) that were deposited in alternating layers with blends of poly(allylamine hydrochloride) (PAH), PVA, and PAA. Although mixing cationic PAH and anionic PAA solutions generally causes polyelectrolyte−polyelectrolyte complexes (PECs) to form through electrostatic interactions, we found that PAH and PAA hardly formed PECs when PVA was present in the solution containing PAA. Consequently, PAA behaved differently in cationic and anionic solutions, resulting in the formation of coatings with hierarchical texture. The structures possessed antireflective properties with a graded refractive index and >95% transmittance. The coatings also displayed resistance to protein adsorption derived from free hydroxyl groups and antifogging performance caused by hydrophilicity combined with the strong hydrogen bonding ability of PVA. The results of this study would be valuable for the development of innovative biomedical devices through a simple and environmentally friendly approach.
■
INTRODUCTION
Among the properties required for a biomaterial surface, antithrombogenicity is crucial for providing long-term bloodcontacting ability.5,6 The in vivo contact of blood with synthetic surfaces like medical devices leads to immediate fibrinogen adsorption and platelet adhesion, which mediates subsequent biological processes such as activation leading to thrombosis and microbial infections.7 To decrease the extent of protein fouling, previous studies have developed hydrophilic surfaces with neutral or balanced charge groups, chain flexibility, and bound water, such as polyethylene glycol (PEG),8 poly(2methoxyethyl acrylate),9 and zwitterionic polymers and their block copolymers,10,11 by mimicking the highly hydrated structure of the cellular glycocalyx.12,13 Similarly, heparin films have been used to prevent coagulation, showing desirable outcomes with regard to activation and inflammation of blood
Over the past decade, the number of endoscopic surgeries such as thoracoscopy and laparoscopy has sharply increased because these low-risk operations promote early recovery of patients.1 Such minimally invasive approaches have become widely used in various clinical departments, including digestive surgery, thoracic surgery, urology, and gynecology. Although laparoscopic surgery is conducted by visualization through an endoscope lens, low visibility in the images caused by protein fouling and/or fog formation on the lens surface remains a critical issue.2,3 The lens needs to be cleaned during each operation to remove biological fluids, protein, fog, and blood, resulting in increased medical expenses, patient cost, and surgical time, along with other problems including the potential risk of blood clots and the introduction of foreign material. Therefore, there is a critical need for functional coatings with high transparency that can prevent the formation of blood clots and fog on endoscope lenses.4 © 2017 American Chemical Society
Received: February 6, 2017 Revised: May 18, 2017 Published: May 18, 2017 4745
DOI: 10.1021/acs.chemmater.7b00465 Chem. Mater. 2017, 29, 4745−4753
Article
Chemistry of Materials humoral factors.14,15 Inspired by these previous studies, we designed antithrombogenic coatings containing poly(allylamine hydrochloride) (PAH), poly(vinyl alcohol) (PVA), and poly(acrylic acid) (PAA).16 Despite considerable effort, the practical use of some medical devices remains challenging because of the issues of fog formation and insufficient transparency mentioned above. In the case of endoscopic surgery, the lens is exposed to a high-humidity environment [>90% relative humidity (RH)] and the difference in temperature between the operating theater and human body results in a dramatic decrease in imaging capability induced by the formation of fog on the lens.17,18 To solve this problem, researchers have concentrated on improving the antifogging performance of many kinds of surfaces.19−21 One effective approach to alleviate fogging is applying a hydrophilic coating to a surface.22 Coatings containing hydrophilic polymers, such as PEG, PVA, and hyaluronic acid, can spread to form a continuous water membrane, and their strong hydrogen bonding interactions with water molecules can promote absorption of condensed water droplets.23−25 We focused our attention on the recent advances in antifogging surfaces that are similar to the surface modification strategies used to realize antithrombogenicity. Moreover, the fabrication of antifogging surfaces via layer-by-layer (LbL) selfassembly is attractive because such biocompatible, economical, and environmentally friendly materials and methods can accelerate the development of multifunctional medical coatings.26−29 LbL self-assembly allows the surface structure and thickness of a film to be controlled on micro- and nanoscales.30−32 A range of materials that are negatively or positively charged, such as monomers, polyelectrolytes, and nanoparticles, have been used in LbL self-assembly.33−37 Typically, films have been prepared by alternate deposition of cationic and anionic polyelectrolytes, each containing a single species. Rubner and co-workers reported that films of PAH and PAA formed by LbL self-assembly possess a unique surface texture under certain conditions.38 Mendelsohn et al. explained the formation process of the surface texture and demonstrated the cytophilicity and cytophobicity of PAH−PAA films.39 Other studies applied polyelectrolyte−polyelectrolyte complexes (PECs) obtained by combining anionic and cationic polymers to drug delivery, water purification, and cosmetics.40−42 Moreover, superhydrophobic and porous coatings and films showing exponential growth have been fabricated through LbL deposition of PECs.43−45 Thus, films produced using LbL selfassembly have been developed for various applications by changing the surface structures based on their intended use.46−51 Here, we fabricate antithrombogenic, antireflective, antifogging surfaces by controlling nano-ordered hierarchical structures produced via LbL self-assembly. Films with hierarchical texture are prepared from PAH−PVA−PAA blends (as the cationic solution) alternately deposited with PVA−PAA blends (as the anionic solution). Usually, mixing PAA with PAH will form PECs (Scheme 1). When PAA chains interact with PAH chains through electrostatic attraction to form PECs, the PAA chains lose the ability to contract freely and cannot form a texture derived from the contraction of PAA chains. However, we discovered that blends of PAA and PAH hardly form PECs when a mixture of PVA and PAA (denoted PVA− PAA) is added to PAH; this prevents PAA from interacting
Scheme 1. Diagrams of Polyelectrolyte−Polyelectrolyte Complex (PEC) Formation (top) and Prevention of PEC Formation for Obtaining Hierarchical Texture via LbL SelfAssembly (bottom)
with PAH, allowing the formation of a PAA-derived texture in the cationic layer (Scheme 1). The hierarchical texture in the film is derived from the different interaction of PAA in the cationic and anionic layers caused by the coexistence of noninteracting PAA and PAH without the aggregation and the hydrogen bonding between PAA and PVA, respectively. The hierarchical texture is more porous and has an aspect ratio higher than those of PAH/PAA and PAH/(PVA−PAA) films, and the films also possess antireflective properties because of a gradual change in the refractive index from top to bottom. The (PAH−PVA−PAA)/(PVA−PAA) film with the highest transmittance is evaluated as an antithrombogenic film by measuring the peak area of fibrinogen adsorbed from an aqueous solution. This film exhibits antifogging performance because of its free hydroxyl (OH) groups and hydrophilicity with the strong hydrogen bonds of PVA. These multifunctional coatings that do not depend on the shape or size of the substrate may facilitate the development of innovative medical devices that display excellent antithrombogenicity, antireflection, antifogging, and biocompatibility.
■
RESULTS AND DISCUSSION Surface Morphologies. The surface of the (PAH/(PVA− PAA))10 film showed ridges that were ∼150 nm wide, similar to that of the PAH/PAA films (Figure 1). The observed texture was caused by the interaction between carboxyl acid (COOH) groups of PAA and amino (NH2) groups of PAH.52 The ((PAH−PAA)/(PVA−PAA))10 film possessed a texture similar to that of the (PAH/(PVA−PAA))10 one with PAH−PAA PECs apparent on its surface. In contrast, the ((PAH−PVA− PAA)/(PVA−PAA))10 film possessed hierarchical texture with large pores between the ridges. Analysis of PAH, PAH−PAA, and PAH−PVA−PAA Solutions. Panels a−d of Figure 2 show photographs of each polyelectrolyte blend solution before and after pH adjustment. The average diameter of the PECs in each solution is displayed under the corresponding photograph (Figure S1). Contraction of PAA according to the change in the degree of dissociation of PAA is the main factor controlling the texture of the films. Some COOH groups of PAA dissociate at pH 3.5.53 Consequently, when a solution containing only PAA was added to a solution of PAH, PAA formed PECs immediately with PAH, and the PAH−PAA solution became cloudy (Figure 2a). When the pH of this PAH−PAA solution was adjusted to 7.5, 4746
DOI: 10.1021/acs.chemmater.7b00465 Chem. Mater. 2017, 29, 4745−4753
Article
Chemistry of Materials
Figure 1. Scanning electron microscopy images of (PAH/(PVA−PAA))10 (left), ((PAH−PAA)/(PVA−PAA))10 (center), and ((PAH−PVA− PAA)/(PVA−PAA))10 (right) films.
Figure 2. Photograph of (a) a PAH−PAA solution at pH 3.5, (b) a PAH−PAA solution at pH 7.5, (c) a PAH−PVA−PAA solution at pH 3.5, and (d) a PAH−PVA−PAA solution at pH 7.5. The corresponding average diameter of the PECs is given below (Figure S1). (e) Fourier transform infrared spectra of each polymer solution at different pH values. (f) Schematic illustration and atomic force microscopy (AFM) images of PAH/ (PVA−PAA), (PAH−PAA)/(PVA−PAA), and (PAH−PVA−PAA)/(PVA−PAA) films. AFM images of LbL films with 9.5 and 10 bilayers are shown on the left and right, respectively. The scan size (x and y) is 2 μm, and the data scale (z) is 200 nm.
the degree of dissociation of COOH groups in the solution increased (Scheme 1).53 As a result, more and larger PECs formed, generating a sizable aggregate (Figure 2b). In contrast, when a PVA−PAA solution at pH 3.5 was added to a solution of PAH, PECs hardly formed (Figure 2c). In this case, COOH groups of PAA formed hydrogen bonds with OH groups of PVA, so PVA surrounded PAA. Thus, PVA prevented PEC formation. Even when the pH of the PAH−PVA−PAA solution was adjusted to 7.5, the blend solution remained almost clear because PAA−PAH PECs hardly formed in the presence of PVA (Figure 2d). Under these conditions, it
appears that there were absolute numbers of negatively charged COO− groups and positively charged NH3+ groups. In other words, PAH and PAA coexisted without the aggregations derived from the electrostatic interaction of NH2 and COOH groups. To support the discussion given above, the dissociation of polymers in each solution at different pH was investigated (Figure 2e). For a solution of only PAA, the COOH groups of PAA began to dissociate rapidly when the pH of the solution reached 6.5. For PVA−PAA and PAH−PVA−PAA solutions, the COOH groups of PAA began to dissociate slightly when 4747
DOI: 10.1021/acs.chemmater.7b00465 Chem. Mater. 2017, 29, 4745−4753
Article
Chemistry of Materials the pH of the solution was 6.5. This difference is caused by hydrogen bonds that formed between the OH groups of PVA and COOH groups of PAA, which gradually collapsed as the pH increased. The COOH groups of PAA were hardly dissociated by mixing with PVA and did not form PECs with PAH. Consequently, PAA in PAH solution could be elasticized and form a wormlike structure. Formation of Hierarchical Texture. When a solution of PAH was used as the cationic layer, the surface was relatively flat when the top layer was PAH (Figure 2f and Figure S3). The average height of the ridges was ∼43.0 nm. This is because PAA, which is the main component that forms texture, was not the top layer. When the next anionic layer, PVA−PAA, was adsorbed on the PAH layer, texture formed on the surface. The average height of the texture from top to bottom was ∼97.2 nm. In the case in which a PAH−PAA solution was used as the cationic layer, some PAH−PAA PECs were present on the flat surface when the top layer was PAH−PAA (Figures 1 and 2f). The average height of the ridges outside the PAH−PAA aggregates was ∼42.4 nm. PAA was present but unable to contract and form texture because of its interaction with PAH. When the next anionic layer, PVA−PAA, was adsorbed on the PAH layer, PAA formed a textured surface. In this film, PAH− PAA PECs with a single texture appeared on the film surface. The average height of the texture outside the PECs was ∼99.2 nm from top to bottom. For (PAH−PVA−PAA)/(PVA−PAA) films, some texture appeared when the top layer was PAH−PVA−PAA (Figure 2f and Figure S3). Atomic force microscopy (AFM) crosssectional analysis confirmed that the roughness of this surface was higher than those of PAH and PAH−PAA layers. The average height of the texture was ∼70.3 nm. In these films, PAA in the PAH−PVA−PAA layer did not interact with PAH, so it was able to contract and form texture in the cationic layer. Moreover, when the next anionic layer (PVA−PAA) was adsorbed, the PAA in this layer formed a different texture on the underlying textured surface to provide a surface with hierarchical texture. The average height of the hierarchical texture from top to bottom exceeded 152 nm. The aspect ratio of the ((PAH−PVA−PAA)/(PVA−PAA))10 surface was higher than that of the (PAH/PAA)21 surface (Figure S4). Optical Properties of the Hierarchical Textured Surfaces. The ((PAH−PVA−PAA)/(PVA−PAA))10 films with hierarchical texture were porous and possessed an aspect ratio higher than that of (PAH/(PVA−PAA))10 films. In general, porous films have a low refractive index, which decreases optical reflection.54 Moreover, structures with a high aspect ratio also exhibit low optical reflection because the refractive index changes gradually.55,56 Accordingly, the film with hierarchical texture exhibited antireflective properties. Figure 3 shows the transmittance of LbL films with different cationic layers. The transmittance of the (PAH/(PVA− PAA))10 film increased slightly from that of a bare glass substrate. The transmittance of the ((PAH−PAA)/(PVA− PAA))10 film was much lower than that of a bare glass substrate. It is believed that this effect was caused by the presence of PAH−PAA PECs inside and on the surface of the film. In contrast, the ((PAH−PVA−PAA)/(PVA−PAA))10 film with hierarchical texture exhibited high transmittance. Its transmittance was >95% from 650 to 850 nm, indicating that the hierarchical texture endowed the film with strong antireflective ability.
Figure 3. Ultraviolet−visible transmittance spectra of (PAH/(PVA− PAA))10, ((PAH−PAA)/(PVA−PAA))10, and ((PAH−PVA−PAA)/ (PVA−PAA))10 films and a glass substrate.
Next, the reason why the hierarchical texture produced antireflective activity is discussed. Panels a−d of Figure 4 show scanning electron microscopy (SEM) images of (PAH−PVA− PAA)/(PVA−PAA) films with different numbers of layers. For the film with eight bilayers, a texture similar to that of the (PAH/(PVA−PAA))10 film was observed. The film containing nine bilayers showed texture with increased height but was not hierarchical. For films with 10 and 11 bilayers, the texture was hierarchical. The formation process of this hierarchical texture is illustrated in Figure 2f. Panels e and f of Figure 4 show the simulated change in the refractive index for these films with 8−11 bilayers on glass substrates. The simulated change in refractive index was calculated using eq 157,58 n2 = (1 − fpore )n polymer 2 + fpore nair 2
(1)
where npolymer, nair, and f pore are the refractive indices of the polymer and air and the pore fraction of the film, respectively. Second, the pore ratio at a certain height was calculated by AFM bearing analysis, which allowed the height distribution of the films to be determined (Figure S5). The value for which the height distribution up to a certain height is integrated is equal to the pore ratio of that height. (Note that actual pore ratios are lower than the values calculated by AFM because the AFM probe cannot scan the pores inside the film.) The refractive index of a cast (PAH−PVA−PAA)/(PVA−PAA) film was 1.49, which was considered to be npolymer. The simulation of the films with 8−11 bilayers confirmed that their refractive indices gradually increased from the top to the bottom of the films. In addition, the height of the hierarchical texture increased gradually as the number of layers increased. For ideal antireflection, a high aspect ratio and a gradual change in the refractive index are required. In the film with 11 bilayers, the height of the texture was approximately 280 nm, and the change in refractive index was the most gradual of those of the films considered. Panels g and h of Figure 4 illustrate the transmittance of films with different numbers of bilayers on glass substrates and a photograph of coated and uncoated glass substrates, respectively. As the number of bilayers increased, transmittance increased and the maximal transmittance shifted to longer wavelengths.37 Transmittance increased as the height of the hierarchical texture increased. The low transmittance observed at short wavelengths was caused by light scattering by the hierarchical texture. 4748
DOI: 10.1021/acs.chemmater.7b00465 Chem. Mater. 2017, 29, 4745−4753
Article
Chemistry of Materials
Figure 4. SEM images of ((PAH−PVA−PAA)/(PVA−PAA))n films with (a) n = 8, (b) n = 9, (c) n = 10, and (d) n = 11. (e) Schematic of a gradual change in the refractive index. (f) Simulated change in the refractive index with the number of bilayers on glass substrates. (g) Transmittance of films with different numbers of bilayers on glass substrates. (h) Photograph of glass substrates exposed to fluorescent light: (left) a glass substrate with a ((PAH−PVA−PAA)/(PVA−PAA))10 film coating and (right) a bare glass substrate without a film coating.
Figure 5. Fourier transform infrared spectra of (a) a bare silicon wafer substrate and fibrinogen and (b) a ((PAH−PVA−PAA)/(PVA−PAA))10 film before and after contact with a fibrinogen solution. SEM images of (c) a silicon wafer and (d and e) a ((PAH−PVA−PAA)/(PVA−PAA))10 film after contact with a fibrinogen solution. F1s core-level spectra of (f) (PAH/(PVA−PAA))10 and (g) ((PAH−PVA−PAA)/(PVA−PAA))10 films derivatized with trifluoroacetic anhydride (TFAA) and cleaved with aqueous NH3. Blank indicates the films before derivatization with TFAA.
Comparison with a SEM image of fibrinogen on a surface without a coating (Figure 5c) indicated that fibrinogen was not present on the hierarchical film surface. Biocompatibility. OH groups can generate a hydration layer on the surface of the films and inhibit adsorption of fibrinogen.59 The (PAH/PAA)10 film was analyzed by F1s and N1s X-ray photoelectron spectroscopy (XPS).60 F1s spectra were recorded to assess the extent of trifluoroacetic anhydride (TFAA) absorption, while N1s spectra detected the presence of NH2 from PAH in the films. The F1s spectra of (PAH/PAA)10
Protein Adsorption. The appearance of the specific amide I band (1600−1700 cm−1) and amide II band (1500−1560 cm−1) was used to detect the adsorption of fibrinogen on a hierarchical film (Figure 5a).16 The results clearly indicate that fibrinogen cannot adsorb to the surface of the ((PAH−PVA− PAA)/(PVA−PAA))10 film because neither of the amide bands was detected and the spectra were similar before and after immersion in a fibrinogen solution (Figure 5b). Panels d and e of Figure 5 show SEM images of a ((PAH−PVA−PAA)/ (PVA−PAA))10 film after contact with a fibrinogen solution. 4749
DOI: 10.1021/acs.chemmater.7b00465 Chem. Mater. 2017, 29, 4745−4753
Article
Chemistry of Materials
Table 1. Changes in the Surface Atomic Concentration of (PAH/PAA)10, (PAH/(PVA−PAA))10, and ((PAH−PVA−PAA)/ (PVA−PAA))10 Films after Derivatization with Trifluoroacetic Anhydride (TFAA) and Cleavage in Aqueous NH3 (PAH/PAA)10
(PAH/(PVA−PAA))10
((PAH−PVA−PAA)/(PVA−PAA))10
sample
%C
%O
%F
%N
%C
%O
%F
%N
%C
%O
%F
%N
blank derivatized with TFAA cleaved with NH3
70.36 66.71 66.44
20.41 21.61 21.59
− 2.10 2.43
9.23 9.58 9.54
67.79 59.20 65.05
22.78 19.76 21.71
− 13.72 2.90
9.43 7.31 10.33
62.75 49.21 62.32
29.20 18.56 24.23
− 24.65 3.51
8.05 7.59 9.94
are shown in Figure S6a. The blank spectrum did not show an F1s peak. However, a peak appeared after derivatization with TFAA, and it did not change after cleavage with NH3. Therefore, TFAA hardly reacts with NH3, and desorption of TFAA cannot occur, although it reacted slightly on the film. The N1s spectra of (PAH/PAA)10 are presented in Figure S6b. These spectra indicate that few NH2 groups of PAH are exposed on the film surface because the peaks around 405 eV do not change throughout the biocompatibility test. Figure 5f depicts F1s spectra of the (PAH/(PVA−PAA))10 film. No F1s peak was detected for the blank sample. After derivatization with TFAA, the film displayed an F1s peak at 692 eV, indicating adsorption of TFAA to OH or NH2 on the film surface. The intensity of this peak dramatically decreased after cleavage with NH3. TFAA that reacted at NH2 groups cannot desorb, whereas TFAA that reacted at OH groups can desorb by reacting with NH3, like in the derivatization reaction (Scheme S1). The F1s spectra of the ((PAH−PVA−PAA)/(PVA− PAA))10 film are provided in Figure 5g. The blank spectrum did not contain a peak. A peak at 692 eV that is stronger than that of the (PAH/(PVA−PAA))10 film was detected after derivatization with TFAA. Moreover, the intensity of this peak dramatically decreased after cleavage with NH3, similar to the behavior of the (PAH/(PVA−PAA))10 film. The surface atomic concentrations of (PAH/(PVA−PAA))10 and ((PAH−PVA−PAA)/(PVA−PAA))10 films were measured to compare their amounts of OH groups, which ares reflected by the amounts of F after derivatization with TFAA (Table 1). The F ratio was 13.72% in the (PAH/(PVA−PAA))10 film and 24.65% in the ((PAH−PVA−PAA)/(PVA−PAA))10 film. The ((PAH−PVA−PAA)/(PVA−PAA))10 film displayed a relatively large F ratio, in other words, a large OH ratio, because this film had PVA in both cationic and anionic layers and OH groups of PVA appeared on the film surface. These results suggest that the films fabricated in this work have biocompatibilities higher than those in our previous work that demonstrated excellent antithrombogenicity.16 Antifogging Performance. After the film had been cooled in a refrigerator to
